as of right now, the company's value chain is much stronger than its primary competitors
One of the biggest reasons that I am very cautious about BEAS' position is that this market is in its infancy. Java appservers have been around for a few years, but without the J2EE extensions, they were not ready for any serious applications. The J2EE standard is still very young and full implementations younger still. Many, many more companies will be adopting these in the future than have so far and, let's not forget, that conceptually a J2EE server is an appliance, i.e., just because you developed on brand X doesn't mean that you can't deploy on brand Y. In practice, it isn't quite that simple, but the penalty for switching is dropping, particularly if you stick to well defined J2EE standards and are careful about exploiting extensions.
IBM's and SUN's dilemma is that the more that companies adopt their respective platforms, the more IBM's and SUN's other prodcuts are marginalized.
I don't know that I agree with this. There is some element of competitiveness in IBM's offerings, maybe, but I sure don't see it with SUN. Everything they have that matters is now in iPlanet ... the Fortran and C++ tools are not a meaningful part of revenue.
Paul Philp would likely argue that one reason its value chain is so large is because of the number of components built by BEAS that make it relatively easy (less time consuming and, thus, less expensive) to write software using the BEAS platform. He has in fact presented a compelling argument that those components constitute BEAS's primary barrier to entry.
Again, there is a timing factor here, but I sure don't see BEAS having anywhere near the kind of rich offering that iPlanet has, particularly when they release FJEE this summer.
What other threat could overtake BEAS?
One of the factors which is more significant than it otherwise would be is when these servers hit the market. People were chomping at the bit wanting something to work on so first out of the gate picked up sales ... but, this may or may not stick depending on how the value proposition shifts.
My current thinking: Microsoft's .NET initiative. Nuff said.
.NET is not really directly competitive. When it grows up enough to turn into something, .NET will need an appserver too.
Does anyone have information showing clearly whether or not there has been or is an ongoing tornado? Based on BEAS's revenue, I don't see a tornado unless I missed an important detail breaking out product segments in the 10k. I haven't done any research into Sun, Oracle or IBM's app server revenue to see what that data offers. On the other hand, how could there not be a tornado if Giga Information is right about market growth to $9 billion in a few years?
One needs to remember that volume comes with deployment, first you have to develop. I would expect a year or more lag between initial acquisition for development and volume deployment of any substantial application. Plus, it is the risk takers who have dived into J2EE applications first. As they are successful, more cautious types will jump on. It may not be a deep chasm, but there is a bit of a chasm to cross here nevertheless. Of particular concern at the high end are performance issues since Java is still more than 2X slower than corresponding C++ at this point. |
